| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.554016620498615, | |
| "eval_steps": 500, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002770083102493075, | |
| "grad_norm": 15.603601455688477, | |
| "learning_rate": 7.5e-06, | |
| "loss": 1.4608, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.00554016620498615, | |
| "grad_norm": 14.760490417480469, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.428, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008310249307479225, | |
| "grad_norm": 19.810142517089844, | |
| "learning_rate": 2.25e-05, | |
| "loss": 1.2831, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0110803324099723, | |
| "grad_norm": 6.9140729904174805, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0688, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.013850415512465374, | |
| "grad_norm": 6.4214253425598145, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.2067, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01662049861495845, | |
| "grad_norm": 11.87337589263916, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.4816, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.019390581717451522, | |
| "grad_norm": 6.669366359710693, | |
| "learning_rate": 5.25e-05, | |
| "loss": 1.1287, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.0221606648199446, | |
| "grad_norm": 8.616178512573242, | |
| "learning_rate": 6e-05, | |
| "loss": 1.2203, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.024930747922437674, | |
| "grad_norm": 9.265680313110352, | |
| "learning_rate": 5.999881193903433e-05, | |
| "loss": 1.1897, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.027700831024930747, | |
| "grad_norm": 5.8777852058410645, | |
| "learning_rate": 5.999524785023657e-05, | |
| "loss": 1.0788, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.030470914127423823, | |
| "grad_norm": 5.538949012756348, | |
| "learning_rate": 5.998930801589704e-05, | |
| "loss": 1.0736, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.0332409972299169, | |
| "grad_norm": 5.686643123626709, | |
| "learning_rate": 5.9980992906474764e-05, | |
| "loss": 1.1573, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.036011080332409975, | |
| "grad_norm": 6.211599349975586, | |
| "learning_rate": 5.9970303180560196e-05, | |
| "loss": 1.1618, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.038781163434903045, | |
| "grad_norm": 5.079687595367432, | |
| "learning_rate": 5.995723968482309e-05, | |
| "loss": 0.9948, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04155124653739612, | |
| "grad_norm": 5.150979042053223, | |
| "learning_rate": 5.994180345394539e-05, | |
| "loss": 1.0328, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.0443213296398892, | |
| "grad_norm": 5.891695499420166, | |
| "learning_rate": 5.9923995710539324e-05, | |
| "loss": 1.1854, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04709141274238227, | |
| "grad_norm": 5.33378791809082, | |
| "learning_rate": 5.990381786505055e-05, | |
| "loss": 1.0424, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.04986149584487535, | |
| "grad_norm": 4.58151912689209, | |
| "learning_rate": 5.988127151564644e-05, | |
| "loss": 1.0392, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 5.48449182510376, | |
| "learning_rate": 5.9856358448089506e-05, | |
| "loss": 1.147, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.055401662049861494, | |
| "grad_norm": 5.061849117279053, | |
| "learning_rate": 5.9829080635595944e-05, | |
| "loss": 1.1072, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05817174515235457, | |
| "grad_norm": 4.828271389007568, | |
| "learning_rate": 5.979944023867938e-05, | |
| "loss": 1.1132, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.060941828254847646, | |
| "grad_norm": 5.035257816314697, | |
| "learning_rate": 5.976743960497973e-05, | |
| "loss": 1.0272, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.06371191135734072, | |
| "grad_norm": 5.55271577835083, | |
| "learning_rate": 5.973308126907723e-05, | |
| "loss": 1.1674, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.0664819944598338, | |
| "grad_norm": 4.926880836486816, | |
| "learning_rate": 5.969636795229172e-05, | |
| "loss": 1.0765, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06925207756232687, | |
| "grad_norm": 5.437366962432861, | |
| "learning_rate": 5.965730256246713e-05, | |
| "loss": 1.1344, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.07202216066481995, | |
| "grad_norm": 4.4522786140441895, | |
| "learning_rate": 5.9615888193741094e-05, | |
| "loss": 0.9976, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.07479224376731301, | |
| "grad_norm": 5.204329967498779, | |
| "learning_rate": 5.9572128126299946e-05, | |
| "loss": 1.1298, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.07756232686980609, | |
| "grad_norm": 5.264033317565918, | |
| "learning_rate": 5.9526025826118885e-05, | |
| "loss": 1.0428, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08033240997229917, | |
| "grad_norm": 5.672830581665039, | |
| "learning_rate": 5.947758494468746e-05, | |
| "loss": 1.1636, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.08310249307479224, | |
| "grad_norm": 5.798141002655029, | |
| "learning_rate": 5.942680931872036e-05, | |
| "loss": 1.1112, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08587257617728532, | |
| "grad_norm": 4.760410308837891, | |
| "learning_rate": 5.937370296985354e-05, | |
| "loss": 1.0444, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.0886426592797784, | |
| "grad_norm": 5.4676313400268555, | |
| "learning_rate": 5.931827010432566e-05, | |
| "loss": 1.0957, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.09141274238227147, | |
| "grad_norm": 4.983479976654053, | |
| "learning_rate": 5.9260515112644995e-05, | |
| "loss": 1.0734, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.09418282548476455, | |
| "grad_norm": 5.242843151092529, | |
| "learning_rate": 5.9200442569241606e-05, | |
| "loss": 1.0879, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.09695290858725762, | |
| "grad_norm": 5.12119722366333, | |
| "learning_rate": 5.9138057232105084e-05, | |
| "loss": 1.0841, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.0997229916897507, | |
| "grad_norm": 5.087031841278076, | |
| "learning_rate": 5.9073364042407705e-05, | |
| "loss": 1.0343, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.10249307479224377, | |
| "grad_norm": 5.226377487182617, | |
| "learning_rate": 5.900636812411301e-05, | |
| "loss": 1.0775, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 5.388728141784668, | |
| "learning_rate": 5.893707478357005e-05, | |
| "loss": 1.0439, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.10803324099722991, | |
| "grad_norm": 5.6382951736450195, | |
| "learning_rate": 5.886548950909301e-05, | |
| "loss": 1.096, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.11080332409972299, | |
| "grad_norm": 5.046523094177246, | |
| "learning_rate": 5.879161797052658e-05, | |
| "loss": 0.9928, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.11357340720221606, | |
| "grad_norm": 5.98545503616333, | |
| "learning_rate": 5.8715466018796865e-05, | |
| "loss": 1.1639, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.11634349030470914, | |
| "grad_norm": 5.064982891082764, | |
| "learning_rate": 5.863703968544795e-05, | |
| "loss": 0.9878, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.11911357340720222, | |
| "grad_norm": 5.36715841293335, | |
| "learning_rate": 5.8556345182164176e-05, | |
| "loss": 1.056, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.12188365650969529, | |
| "grad_norm": 4.7610368728637695, | |
| "learning_rate": 5.8473388900278206e-05, | |
| "loss": 0.9625, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.12465373961218837, | |
| "grad_norm": 5.5188889503479, | |
| "learning_rate": 5.83881774102647e-05, | |
| "loss": 1.0828, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.12742382271468145, | |
| "grad_norm": 5.703125476837158, | |
| "learning_rate": 5.8300717461220027e-05, | |
| "loss": 1.0632, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.13019390581717452, | |
| "grad_norm": 5.264132499694824, | |
| "learning_rate": 5.821101598032759e-05, | |
| "loss": 1.0057, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1329639889196676, | |
| "grad_norm": 6.05692720413208, | |
| "learning_rate": 5.811908007230929e-05, | |
| "loss": 1.0565, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.13573407202216067, | |
| "grad_norm": 4.773650169372559, | |
| "learning_rate": 5.802491701886268e-05, | |
| "loss": 0.9764, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.13850415512465375, | |
| "grad_norm": 5.461302757263184, | |
| "learning_rate": 5.792853427808431e-05, | |
| "loss": 1.0002, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.14127423822714683, | |
| "grad_norm": 5.83071231842041, | |
| "learning_rate": 5.7829939483878996e-05, | |
| "loss": 1.0353, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1440443213296399, | |
| "grad_norm": 5.715704917907715, | |
| "learning_rate": 5.772914044535516e-05, | |
| "loss": 0.985, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.14681440443213298, | |
| "grad_norm": 5.309004306793213, | |
| "learning_rate": 5.762614514620634e-05, | |
| "loss": 1.0106, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.14958448753462603, | |
| "grad_norm": 5.693915843963623, | |
| "learning_rate": 5.752096174407884e-05, | |
| "loss": 1.0049, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1523545706371191, | |
| "grad_norm": 5.644995212554932, | |
| "learning_rate": 5.741359856992561e-05, | |
| "loss": 1.0411, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.15512465373961218, | |
| "grad_norm": 5.765170574188232, | |
| "learning_rate": 5.73040641273464e-05, | |
| "loss": 1.0039, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 5.643161296844482, | |
| "learning_rate": 5.719236709191428e-05, | |
| "loss": 1.0173, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.16066481994459833, | |
| "grad_norm": 5.821686267852783, | |
| "learning_rate": 5.707851631048841e-05, | |
| "loss": 0.9725, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.1634349030470914, | |
| "grad_norm": 6.034667015075684, | |
| "learning_rate": 5.6962520800513414e-05, | |
| "loss": 1.0381, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.16620498614958448, | |
| "grad_norm": 5.336913585662842, | |
| "learning_rate": 5.6844389749305136e-05, | |
| "loss": 0.9266, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.16897506925207756, | |
| "grad_norm": 5.65543270111084, | |
| "learning_rate": 5.672413251332297e-05, | |
| "loss": 0.9768, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.17174515235457063, | |
| "grad_norm": 5.3824615478515625, | |
| "learning_rate": 5.6601758617428766e-05, | |
| "loss": 0.9276, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1745152354570637, | |
| "grad_norm": 5.195915222167969, | |
| "learning_rate": 5.647727775413245e-05, | |
| "loss": 0.9332, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.1772853185595568, | |
| "grad_norm": 5.240732669830322, | |
| "learning_rate": 5.6350699782824346e-05, | |
| "loss": 0.9012, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.18005540166204986, | |
| "grad_norm": 5.112541675567627, | |
| "learning_rate": 5.622203472899423e-05, | |
| "loss": 0.9211, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.18282548476454294, | |
| "grad_norm": 6.605805397033691, | |
| "learning_rate": 5.609129278343731e-05, | |
| "loss": 0.9847, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.18559556786703602, | |
| "grad_norm": 5.1862568855285645, | |
| "learning_rate": 5.595848430144705e-05, | |
| "loss": 0.9064, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1883656509695291, | |
| "grad_norm": 6.03863525390625, | |
| "learning_rate": 5.582361980199504e-05, | |
| "loss": 0.9405, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.19113573407202217, | |
| "grad_norm": 6.4853901863098145, | |
| "learning_rate": 5.568670996689773e-05, | |
| "loss": 0.9676, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.19390581717451524, | |
| "grad_norm": 5.415260314941406, | |
| "learning_rate": 5.554776563997056e-05, | |
| "loss": 0.8873, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.19667590027700832, | |
| "grad_norm": 5.11320686340332, | |
| "learning_rate": 5.540679782616892e-05, | |
| "loss": 0.8752, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.1994459833795014, | |
| "grad_norm": 4.8276214599609375, | |
| "learning_rate": 5.52638176907166e-05, | |
| "loss": 0.8441, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.20221606648199447, | |
| "grad_norm": 5.574111461639404, | |
| "learning_rate": 5.5118836558221475e-05, | |
| "loss": 0.907, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.20498614958448755, | |
| "grad_norm": 5.575868606567383, | |
| "learning_rate": 5.497186591177849e-05, | |
| "loss": 0.8691, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.2077562326869806, | |
| "grad_norm": 5.388404846191406, | |
| "learning_rate": 5.4822917392060184e-05, | |
| "loss": 0.8935, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 6.2548747062683105, | |
| "learning_rate": 5.46720027963947e-05, | |
| "loss": 0.8814, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.21329639889196675, | |
| "grad_norm": 5.358329772949219, | |
| "learning_rate": 5.4519134077831395e-05, | |
| "loss": 0.9041, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.21606648199445982, | |
| "grad_norm": 6.0121564865112305, | |
| "learning_rate": 5.4364323344194095e-05, | |
| "loss": 0.8568, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.2188365650969529, | |
| "grad_norm": 6.2294020652771, | |
| "learning_rate": 5.420758285712211e-05, | |
| "loss": 0.9205, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.22160664819944598, | |
| "grad_norm": 5.889683723449707, | |
| "learning_rate": 5.404892503109906e-05, | |
| "loss": 0.8264, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.22437673130193905, | |
| "grad_norm": 5.401793956756592, | |
| "learning_rate": 5.388836243246963e-05, | |
| "loss": 0.8574, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.22714681440443213, | |
| "grad_norm": 6.194601058959961, | |
| "learning_rate": 5.372590777844421e-05, | |
| "loss": 0.8742, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2299168975069252, | |
| "grad_norm": 5.504939079284668, | |
| "learning_rate": 5.356157393609167e-05, | |
| "loss": 0.8261, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.23268698060941828, | |
| "grad_norm": 5.818756103515625, | |
| "learning_rate": 5.339537392132025e-05, | |
| "loss": 0.8364, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.23545706371191136, | |
| "grad_norm": 5.8196024894714355, | |
| "learning_rate": 5.322732089784661e-05, | |
| "loss": 0.8577, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.23822714681440443, | |
| "grad_norm": 6.102773189544678, | |
| "learning_rate": 5.305742817615325e-05, | |
| "loss": 0.8708, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2409972299168975, | |
| "grad_norm": 6.225861549377441, | |
| "learning_rate": 5.288570921243423e-05, | |
| "loss": 0.8654, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.24376731301939059, | |
| "grad_norm": 6.0107340812683105, | |
| "learning_rate": 5.2712177607529405e-05, | |
| "loss": 0.8401, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.24653739612188366, | |
| "grad_norm": 6.3431715965271, | |
| "learning_rate": 5.2536847105847185e-05, | |
| "loss": 0.855, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.24930747922437674, | |
| "grad_norm": 5.196923732757568, | |
| "learning_rate": 5.235973159427591e-05, | |
| "loss": 0.756, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2520775623268698, | |
| "grad_norm": 6.157230377197266, | |
| "learning_rate": 5.218084510108397e-05, | |
| "loss": 0.856, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.2548476454293629, | |
| "grad_norm": 5.813492774963379, | |
| "learning_rate": 5.200020179480868e-05, | |
| "loss": 0.7829, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.25761772853185594, | |
| "grad_norm": 6.468319416046143, | |
| "learning_rate": 5.181781598313409e-05, | |
| "loss": 0.8327, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.26038781163434904, | |
| "grad_norm": 6.1590423583984375, | |
| "learning_rate": 5.16337021117578e-05, | |
| "loss": 0.8035, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 6.378724575042725, | |
| "learning_rate": 5.14478747632467e-05, | |
| "loss": 0.8134, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2659279778393352, | |
| "grad_norm": 5.47311544418335, | |
| "learning_rate": 5.126034865588208e-05, | |
| "loss": 0.7149, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.26869806094182824, | |
| "grad_norm": 5.790994644165039, | |
| "learning_rate": 5.107113864249381e-05, | |
| "loss": 0.7714, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.27146814404432135, | |
| "grad_norm": 5.880769729614258, | |
| "learning_rate": 5.088025970928399e-05, | |
| "loss": 0.7519, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2742382271468144, | |
| "grad_norm": 5.894231796264648, | |
| "learning_rate": 5.068772697463992e-05, | |
| "loss": 0.8124, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.2770083102493075, | |
| "grad_norm": 6.011329174041748, | |
| "learning_rate": 5.0493555687936704e-05, | |
| "loss": 0.7536, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.27977839335180055, | |
| "grad_norm": 5.813382148742676, | |
| "learning_rate": 5.029776122832945e-05, | |
| "loss": 0.7483, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.28254847645429365, | |
| "grad_norm": 6.445542335510254, | |
| "learning_rate": 5.0100359103535134e-05, | |
| "loss": 0.7376, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2853185595567867, | |
| "grad_norm": 5.571536540985107, | |
| "learning_rate": 4.9901364948604366e-05, | |
| "loss": 0.7538, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.2880886426592798, | |
| "grad_norm": 5.566730499267578, | |
| "learning_rate": 4.970079452468298e-05, | |
| "loss": 0.7102, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.29085872576177285, | |
| "grad_norm": 6.336560249328613, | |
| "learning_rate": 4.949866371776376e-05, | |
| "loss": 0.7469, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.29362880886426596, | |
| "grad_norm": 5.440810203552246, | |
| "learning_rate": 4.929498853742816e-05, | |
| "loss": 0.7252, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.296398891966759, | |
| "grad_norm": 6.162292957305908, | |
| "learning_rate": 4.908978511557827e-05, | |
| "loss": 0.7569, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.29916897506925205, | |
| "grad_norm": 6.059966087341309, | |
| "learning_rate": 4.8883069705159104e-05, | |
| "loss": 0.7136, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.30193905817174516, | |
| "grad_norm": 6.249420642852783, | |
| "learning_rate": 4.867485867887136e-05, | |
| "loss": 0.7408, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3047091412742382, | |
| "grad_norm": 5.173600673675537, | |
| "learning_rate": 4.846516852787457e-05, | |
| "loss": 0.7106, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3074792243767313, | |
| "grad_norm": 5.633162975311279, | |
| "learning_rate": 4.8254015860480934e-05, | |
| "loss": 0.6951, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.31024930747922436, | |
| "grad_norm": 5.940577507019043, | |
| "learning_rate": 4.804141740083993e-05, | |
| "loss": 0.7022, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.31301939058171746, | |
| "grad_norm": 5.466473579406738, | |
| "learning_rate": 4.7827389987613635e-05, | |
| "loss": 0.6694, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 6.02973747253418, | |
| "learning_rate": 4.761195057264307e-05, | |
| "loss": 0.6799, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3185595567867036, | |
| "grad_norm": 5.916580677032471, | |
| "learning_rate": 4.739511621960551e-05, | |
| "loss": 0.7019, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.32132963988919666, | |
| "grad_norm": 5.600930213928223, | |
| "learning_rate": 4.717690410266304e-05, | |
| "loss": 0.6818, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.32409972299168976, | |
| "grad_norm": 5.327589988708496, | |
| "learning_rate": 4.69573315051022e-05, | |
| "loss": 0.6811, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3268698060941828, | |
| "grad_norm": 5.755538463592529, | |
| "learning_rate": 4.673641581796515e-05, | |
| "loss": 0.6705, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3296398891966759, | |
| "grad_norm": 6.217751979827881, | |
| "learning_rate": 4.6514174538672187e-05, | |
| "loss": 0.6856, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.33240997229916897, | |
| "grad_norm": 5.41038179397583, | |
| "learning_rate": 4.629062526963592e-05, | |
| "loss": 0.6509, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.33518005540166207, | |
| "grad_norm": 5.7826247215271, | |
| "learning_rate": 4.606578571686703e-05, | |
| "loss": 0.662, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.3379501385041551, | |
| "grad_norm": 6.125760555267334, | |
| "learning_rate": 4.583967368857194e-05, | |
| "loss": 0.6518, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3407202216066482, | |
| "grad_norm": 5.408969879150391, | |
| "learning_rate": 4.561230709374229e-05, | |
| "loss": 0.6391, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.34349030470914127, | |
| "grad_norm": 5.880269527435303, | |
| "learning_rate": 4.5383703940736494e-05, | |
| "loss": 0.6388, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.3462603878116344, | |
| "grad_norm": 6.065842628479004, | |
| "learning_rate": 4.51538823358534e-05, | |
| "loss": 0.653, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3490304709141274, | |
| "grad_norm": 6.418548107147217, | |
| "learning_rate": 4.49228604818982e-05, | |
| "loss": 0.6607, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3518005540166205, | |
| "grad_norm": 5.7509074211120605, | |
| "learning_rate": 4.469065667674069e-05, | |
| "loss": 0.6289, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.3545706371191136, | |
| "grad_norm": 5.758950710296631, | |
| "learning_rate": 4.445728931186599e-05, | |
| "loss": 0.6332, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.3573407202216066, | |
| "grad_norm": 5.907932758331299, | |
| "learning_rate": 4.42227768709179e-05, | |
| "loss": 0.634, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3601108033240997, | |
| "grad_norm": 5.904737949371338, | |
| "learning_rate": 4.398713792823489e-05, | |
| "loss": 0.5948, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.3628808864265928, | |
| "grad_norm": 6.62101411819458, | |
| "learning_rate": 4.375039114737895e-05, | |
| "loss": 0.6332, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3656509695290859, | |
| "grad_norm": 6.03896951675415, | |
| "learning_rate": 4.3512555279657344e-05, | |
| "loss": 0.6153, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 5.788692474365234, | |
| "learning_rate": 4.3273649162637454e-05, | |
| "loss": 0.6223, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.37119113573407203, | |
| "grad_norm": 5.691258907318115, | |
| "learning_rate": 4.3033691718654767e-05, | |
| "loss": 0.6048, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3739612188365651, | |
| "grad_norm": 5.318130016326904, | |
| "learning_rate": 4.2792701953314085e-05, | |
| "loss": 0.6023, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3767313019390582, | |
| "grad_norm": 6.409132480621338, | |
| "learning_rate": 4.2550698953984335e-05, | |
| "loss": 0.5975, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.37950138504155123, | |
| "grad_norm": 5.915623664855957, | |
| "learning_rate": 4.230770188828665e-05, | |
| "loss": 0.6048, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.38227146814404434, | |
| "grad_norm": 6.316535472869873, | |
| "learning_rate": 4.2063730002576244e-05, | |
| "loss": 0.5944, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.3850415512465374, | |
| "grad_norm": 6.343788146972656, | |
| "learning_rate": 4.181880262041808e-05, | |
| "loss": 0.6115, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.3878116343490305, | |
| "grad_norm": 6.030927658081055, | |
| "learning_rate": 4.15729391410563e-05, | |
| "loss": 0.5709, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39058171745152354, | |
| "grad_norm": 5.820158004760742, | |
| "learning_rate": 4.132615903787774e-05, | |
| "loss": 0.5882, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.39335180055401664, | |
| "grad_norm": 5.914207935333252, | |
| "learning_rate": 4.1078481856869586e-05, | |
| "loss": 0.5579, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3961218836565097, | |
| "grad_norm": 5.729140758514404, | |
| "learning_rate": 4.0829927215071234e-05, | |
| "loss": 0.569, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.3988919667590028, | |
| "grad_norm": 6.345619201660156, | |
| "learning_rate": 4.0580514799020525e-05, | |
| "loss": 0.5574, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.40166204986149584, | |
| "grad_norm": 5.319207668304443, | |
| "learning_rate": 4.033026436319452e-05, | |
| "loss": 0.5584, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.40443213296398894, | |
| "grad_norm": 6.186460018157959, | |
| "learning_rate": 4.0079195728444845e-05, | |
| "loss": 0.5476, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.407202216066482, | |
| "grad_norm": 6.213712692260742, | |
| "learning_rate": 3.982732878042782e-05, | |
| "loss": 0.5403, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4099722991689751, | |
| "grad_norm": 5.564090251922607, | |
| "learning_rate": 3.95746834680294e-05, | |
| "loss": 0.5391, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.41274238227146814, | |
| "grad_norm": 6.060794830322266, | |
| "learning_rate": 3.932127980178519e-05, | |
| "loss": 0.5332, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.4155124653739612, | |
| "grad_norm": 5.621760368347168, | |
| "learning_rate": 3.906713785229546e-05, | |
| "loss": 0.5426, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4182825484764543, | |
| "grad_norm": 5.4746012687683105, | |
| "learning_rate": 3.8812277748635554e-05, | |
| "loss": 0.5444, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 4.968982696533203, | |
| "learning_rate": 3.855671967676153e-05, | |
| "loss": 0.525, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.42382271468144045, | |
| "grad_norm": 5.877557277679443, | |
| "learning_rate": 3.830048387791136e-05, | |
| "loss": 0.5485, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.4265927977839335, | |
| "grad_norm": 5.59464168548584, | |
| "learning_rate": 3.804359064700175e-05, | |
| "loss": 0.5274, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.4293628808864266, | |
| "grad_norm": 5.718829154968262, | |
| "learning_rate": 3.778606033102071e-05, | |
| "loss": 0.5321, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.43213296398891965, | |
| "grad_norm": 5.996484279632568, | |
| "learning_rate": 3.752791332741596e-05, | |
| "loss": 0.5083, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.43490304709141275, | |
| "grad_norm": 5.6276373863220215, | |
| "learning_rate": 3.7269170082479376e-05, | |
| "loss": 0.5466, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.4376731301939058, | |
| "grad_norm": 5.630644798278809, | |
| "learning_rate": 3.7009851089727615e-05, | |
| "loss": 0.5168, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4404432132963989, | |
| "grad_norm": 5.379649639129639, | |
| "learning_rate": 3.6749976888278864e-05, | |
| "loss": 0.529, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.44321329639889195, | |
| "grad_norm": 5.533950328826904, | |
| "learning_rate": 3.6489568061226106e-05, | |
| "loss": 0.4908, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.44598337950138506, | |
| "grad_norm": 6.063986301422119, | |
| "learning_rate": 3.6228645234006835e-05, | |
| "loss": 0.5209, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.4487534626038781, | |
| "grad_norm": 6.18126916885376, | |
| "learning_rate": 3.596722907276946e-05, | |
| "loss": 0.4909, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.4515235457063712, | |
| "grad_norm": 5.693446636199951, | |
| "learning_rate": 3.570534028273645e-05, | |
| "loss": 0.5257, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.45429362880886426, | |
| "grad_norm": 4.973173141479492, | |
| "learning_rate": 3.5442999606564376e-05, | |
| "loss": 0.4936, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.45706371191135736, | |
| "grad_norm": 5.8764729499816895, | |
| "learning_rate": 3.518022782270106e-05, | |
| "loss": 0.4793, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.4598337950138504, | |
| "grad_norm": 4.693322658538818, | |
| "learning_rate": 3.491704574373978e-05, | |
| "loss": 0.4836, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4626038781163435, | |
| "grad_norm": 5.623289108276367, | |
| "learning_rate": 3.465347421477087e-05, | |
| "loss": 0.4814, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.46537396121883656, | |
| "grad_norm": 5.063323497772217, | |
| "learning_rate": 3.438953411173066e-05, | |
| "loss": 0.4801, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.46814404432132967, | |
| "grad_norm": 5.001865863800049, | |
| "learning_rate": 3.412524633974808e-05, | |
| "loss": 0.5035, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4709141274238227, | |
| "grad_norm": 4.878357410430908, | |
| "learning_rate": 3.386063183148883e-05, | |
| "loss": 0.4814, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 6.276041507720947, | |
| "learning_rate": 3.359571154549747e-05, | |
| "loss": 0.4853, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.47645429362880887, | |
| "grad_norm": 5.440047740936279, | |
| "learning_rate": 3.3330506464537356e-05, | |
| "loss": 0.4837, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.4792243767313019, | |
| "grad_norm": 4.922585964202881, | |
| "learning_rate": 3.306503759392882e-05, | |
| "loss": 0.4957, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.481994459833795, | |
| "grad_norm": 4.886894226074219, | |
| "learning_rate": 3.279932595988538e-05, | |
| "loss": 0.4831, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.48476454293628807, | |
| "grad_norm": 4.792039394378662, | |
| "learning_rate": 3.253339260784838e-05, | |
| "loss": 0.5031, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.48753462603878117, | |
| "grad_norm": 4.9557647705078125, | |
| "learning_rate": 3.226725860082018e-05, | |
| "loss": 0.4812, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.4903047091412742, | |
| "grad_norm": 5.699478626251221, | |
| "learning_rate": 3.200094501769581e-05, | |
| "loss": 0.4898, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4930747922437673, | |
| "grad_norm": 4.814623832702637, | |
| "learning_rate": 3.173447295159344e-05, | |
| "loss": 0.4469, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.49584487534626037, | |
| "grad_norm": 5.314681053161621, | |
| "learning_rate": 3.1467863508183735e-05, | |
| "loss": 0.4889, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.4986149584487535, | |
| "grad_norm": 5.161649703979492, | |
| "learning_rate": 3.1201137804018246e-05, | |
| "loss": 0.4543, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.5013850415512465, | |
| "grad_norm": 5.343588829040527, | |
| "learning_rate": 3.093431696485679e-05, | |
| "loss": 0.4557, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.5041551246537396, | |
| "grad_norm": 4.8219499588012695, | |
| "learning_rate": 3.0667422123994304e-05, | |
| "loss": 0.4607, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.5069252077562327, | |
| "grad_norm": 5.002239227294922, | |
| "learning_rate": 3.040047442058694e-05, | |
| "loss": 0.4693, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5096952908587258, | |
| "grad_norm": 4.721946716308594, | |
| "learning_rate": 3.0133494997977792e-05, | |
| "loss": 0.4445, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.5124653739612188, | |
| "grad_norm": 4.8258538246154785, | |
| "learning_rate": 2.9866505002022223e-05, | |
| "loss": 0.4712, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5152354570637119, | |
| "grad_norm": 5.217929840087891, | |
| "learning_rate": 2.9599525579413066e-05, | |
| "loss": 0.4282, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.518005540166205, | |
| "grad_norm": 5.506514072418213, | |
| "learning_rate": 2.9332577876005707e-05, | |
| "loss": 0.4733, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5207756232686981, | |
| "grad_norm": 4.479320049285889, | |
| "learning_rate": 2.9065683035143213e-05, | |
| "loss": 0.4892, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5235457063711911, | |
| "grad_norm": 4.9687089920043945, | |
| "learning_rate": 2.8798862195981763e-05, | |
| "loss": 0.4549, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 5.03638219833374, | |
| "learning_rate": 2.8532136491816266e-05, | |
| "loss": 0.4461, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5290858725761773, | |
| "grad_norm": 5.558093547821045, | |
| "learning_rate": 2.8265527048406564e-05, | |
| "loss": 0.45, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5318559556786704, | |
| "grad_norm": 4.820427417755127, | |
| "learning_rate": 2.7999054982304192e-05, | |
| "loss": 0.4567, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5346260387811634, | |
| "grad_norm": 4.981771945953369, | |
| "learning_rate": 2.7732741399179813e-05, | |
| "loss": 0.4754, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5373961218836565, | |
| "grad_norm": 4.8197245597839355, | |
| "learning_rate": 2.746660739215162e-05, | |
| "loss": 0.4623, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.5401662049861495, | |
| "grad_norm": 4.506078243255615, | |
| "learning_rate": 2.7200674040114627e-05, | |
| "loss": 0.4579, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.5429362880886427, | |
| "grad_norm": 5.280223369598389, | |
| "learning_rate": 2.693496240607119e-05, | |
| "loss": 0.4455, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.5457063711911357, | |
| "grad_norm": 4.943070411682129, | |
| "learning_rate": 2.666949353546265e-05, | |
| "loss": 0.4516, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.5484764542936288, | |
| "grad_norm": 4.689980983734131, | |
| "learning_rate": 2.6404288454502546e-05, | |
| "loss": 0.4353, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.5512465373961218, | |
| "grad_norm": 4.386908531188965, | |
| "learning_rate": 2.6139368168511176e-05, | |
| "loss": 0.47, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.554016620498615, | |
| "grad_norm": 4.444925785064697, | |
| "learning_rate": 2.5874753660251916e-05, | |
| "loss": 0.4538, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 361, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4.5606865011041894e+17, | |
| "train_batch_size": 9, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |